Opto-coupler failure question (HCNR201)

[Joerg]

Richard Rasker wrote:

Joerg wrote:

Richard Rasker wrote:
Jan Panteltje wrote:

On a sunny day (Wed, 09 Sep 2009 20:55:49 +0200) it happened Richard
Rasker <spamtrap@linetec.nl> wrote in
4aa7fa35$0$725$7ade8c0d@textreader.nntp.internl.net>:

Jan Panteltje wrote:

On a sunny day (Wed, 09 Sep 2009 20:02:38 +0200) it happened Richard
Rasker <spamtrap@linetec.nl> wrote in
4aa7edbe$0$730$7ade8c0d@textreader.nntp.internl.net>:

So my question: is this a simple case of "bad luck", or are there
other ways a LED in an opto-isolator may fail in this weird way
(current OK, yet no light)?
Perhaps some part internal to the opto-isolater did break of due to
vibration and blocks the light path?
Did you open the defective one?
No, I didn't (yet) -- but isn't this very, very unlikely?

Richard Rasker
I could also imagine a bind wire coming lose of the photo transistor.
Before doing any destructive investigation, perhaps you could also check
the photo transistor.
Vbe and Vbc diodes, if it has the base on a pin, multimeter + on
collector, - on emittor, and feed the base from the collector via say
100k Ohm. Should draw some current.
The HCNR201 isn't a normal opto-coupler; it doesn't have one photo
transistor, but two photo diodes, one of which is normally used in a
feedback circuit driving the LED. Both photo diodes behave the same, i.e.
they don't respond to any current I send through the LED.

Somehow sounds like a busted LED.

BTW, it's best not to split groups and follow-up fields differently, it
mangles your thread.

Hm, I /did/ set the follow-up to sci.electronics.repair. I don't know what
went wrong ...

No, what I meant was don't post in two NGs and then set the follow-up
only to one. It splits the thread and also leads to double-efforts, like
someone answering while another poster had given the same answer in the
follow-up NG (which he hadn't subscribed to).

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

 

[Joerg]

Richard Rasker wrote:

Spehro Pefhany wrote:

On Wed, 09 Sep 2009 20:55:49 +0200, Richard Rasker
spamtrap@linetec.nl> wrote:

Jan Panteltje wrote:

On a sunny day (Wed, 09 Sep 2009 20:02:38 +0200) it happened Richard
Rasker <spamtrap@linetec.nl> wrote in
4aa7edbe$0$730$7ade8c0d@textreader.nntp.internl.net>:

So my question: is this a simple case of "bad luck", or are there other
ways a LED in an opto-isolator may fail in this weird way (current OK,
yet no light)?
Perhaps some part internal to the opto-isolater did break of due to
vibration and blocks the light path?
Did you open the defective one?
No, I didn't (yet) -- but isn't this very, very unlikely?

Richard Rasker
It's incredibly likely compared to what seems to be the only
alternative- an LED which acts exactly like a AlGaAs D but doesn't LE.

OTOH, an electrically damaged diode that measured something like a
short would not be unusual at all. Could be something like lightning
or RF damage. You don't have the opto in there because it's a benign
environment, eh?

Hehe, spot on -- this is a 24V ship's electrical installation, with heavy DC
motors and other possible sources of interference. That's why I also use
DC-DC-couplers (with a wide-range input) to supply the rest of the
circuitry. That way, input, output and supply are all galvanically
separated.
But the LED still behaves as a LED -- in an electrical sense, that is. It's
not shorted out or anything.

Here is a short article about LED failure modes:

http://www.emsnow.com/cnt/files/White%20Papers/DFRLEDFailures.pdf

I am no expert on this but have done a fair bit of work with laser
diodes, including ones in the >$1k class. It only takes microseconds of
mishap and a $1k laser diode becomes a $1 LED. We also had cases where
the diode looked quite normal electrically but only a miniscule or
absolutely no optical energy was generated by it anymore.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

 

[Richard Rasker]

Joerg wrote:

Richard Rasker wrote:
Spehro Pefhany wrote:

On Wed, 09 Sep 2009 20:55:49 +0200, Richard Rasker
spamtrap@linetec.nl> wrote:

Jan Panteltje wrote:

On a sunny day (Wed, 09 Sep 2009 20:02:38 +0200) it happened Richard
Rasker <spamtrap@linetec.nl> wrote in
4aa7edbe$0$730$7ade8c0d@textreader.nntp.internl.net>:

So my question: is this a simple case of "bad luck", or are there
other ways a LED in an opto-isolator may fail in this weird way
(current OK, yet no light)?
Perhaps some part internal to the opto-isolater did break of due to
vibration and blocks the light path?
Did you open the defective one?
No, I didn't (yet) -- but isn't this very, very unlikely?

Richard Rasker
It's incredibly likely compared to what seems to be the only
alternative- an LED which acts exactly like a AlGaAs D but doesn't LE.

OTOH, an electrically damaged diode that measured something like a
short would not be unusual at all. Could be something like lightning
or RF damage. You don't have the opto in there because it's a benign
environment, eh?

Hehe, spot on -- this is a 24V ship's electrical installation, with heavy
DC motors and other possible sources of interference. That's why I also
use DC-DC-couplers (with a wide-range input) to supply the rest of the
circuitry. That way, input, output and supply are all galvanically
separated.
But the LED still behaves as a LED -- in an electrical sense, that is.
It's not shorted out or anything.


Here is a short article about LED failure modes:

http://www.emsnow.com/cnt/files/White%20Papers/DFRLEDFailures.pdf

I am no expert on this but have done a fair bit of work with laser
diodes, including ones in the >$1k class. It only takes microseconds of
mishap and a $1k laser diode becomes a $1 LED. We also had cases where
the diode looked quite normal electrically but only a miniscule or
absolutely no optical energy was generated by it anymore.

OK, thanks, this is quite interesting. For the time being, I'll just wait
and see if this failure repeats itself (and try to be even more careful
handling and soldering the devices).

Best regards,

Richard Rasker
--
http://www.linetec.nl

 

[Joerg]

Richard Rasker wrote:

Joerg wrote:

Richard Rasker wrote:
Spehro Pefhany wrote:

On Wed, 09 Sep 2009 20:55:49 +0200, Richard Rasker
spamtrap@linetec.nl> wrote:

Jan Panteltje wrote:

On a sunny day (Wed, 09 Sep 2009 20:02:38 +0200) it happened Richard
Rasker <spamtrap@linetec.nl> wrote in
4aa7edbe$0$730$7ade8c0d@textreader.nntp.internl.net>:

So my question: is this a simple case of "bad luck", or are there
other ways a LED in an opto-isolator may fail in this weird way
(current OK, yet no light)?
Perhaps some part internal to the opto-isolater did break of due to
vibration and blocks the light path?
Did you open the defective one?
No, I didn't (yet) -- but isn't this very, very unlikely?

Richard Rasker
It's incredibly likely compared to what seems to be the only
alternative- an LED which acts exactly like a AlGaAs D but doesn't LE.

OTOH, an electrically damaged diode that measured something like a
short would not be unusual at all. Could be something like lightning
or RF damage. You don't have the opto in there because it's a benign
environment, eh?
Hehe, spot on -- this is a 24V ship's electrical installation, with heavy
DC motors and other possible sources of interference. That's why I also
use DC-DC-couplers (with a wide-range input) to supply the rest of the
circuitry. That way, input, output and supply are all galvanically
separated.
But the LED still behaves as a LED -- in an electrical sense, that is.
It's not shorted out or anything.

Here is a short article about LED failure modes:

http://www.emsnow.com/cnt/files/White%20Papers/DFRLEDFailures.pdf

I am no expert on this but have done a fair bit of work with laser
diodes, including ones in the >$1k class. It only takes microseconds of
mishap and a $1k laser diode becomes a $1 LED. We also had cases where
the diode looked quite normal electrically but only a miniscule or
absolutely no optical energy was generated by it anymore.

OK, thanks, this is quite interesting. For the time being, I'll just wait
and see if this failure repeats itself (and try to be even more careful
handling and soldering the devices).

Reminds me of a joke. No, no, this is _not_ meant to apply to your case.

A bowling group returns from a road trip. Coming down a pass the brakes
on the car fade. 40mph ...50 ... 60 ... some guys start to scream. The
driver steers towards the guard rail, lots of sparks fly, some more
screaming, a passenger faints, vehicle scrapes to a stop, all smoking.
Everybody evacuates. One guy, an engineer, looks at the mess: "Tsk, tsk,
tsk, unbelievable. Interesting. Hey, let's take it up back to the top
and see if the failure repeats itself!"

--
SCNR, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

 

[Joerg]

Richard Rasker wrote:

Spehro Pefhany wrote:

On Wed, 09 Sep 2009 20:02:38 +0200, Richard Rasker
spamtrap@linetec.nl> wrote:

Hi all,

I'm using an HCNR201 opto-isolator device in a galvanically isolated
linear signal transfer application.

[snip mysterious breakdown]

So my question: is this a simple case of "bad luck", or are there other
ways a LED in an opto-isolator may fail in this weird way (current OK, yet
no light)?

When you say that the LED behaves like an LED diode, you mean that Vf
is just what you'd expect from figure 9 of the datasheet?

Hmm, I get a Vf of 1.3V @ 10mA If, so that's rather low, but still just
within specifications.

Is the reverse leakage close to typical values?

I don't know what typical valueas are, but at Vr = 1V, I got about 1uA of
leakage current, rising rapidly with higher reverse voltage. At Vr = 3
volts and up, the LED starts conducting whole milli-amps (current-limited
to 2mA) -- so I guess that if the LED wasn't broken to begin with, it is
now.

Whoops, now it probably is. But it probably already was.

But during all this, I monitored both photo diodes, and at no point did they
produce any sigificant voltage into 10MOhm voltmeter inputs.

What are the odds of both photodiodes failing around the same time while
they are connected to very different parts of the circuit? This would
also point to the LED as the culprit, at least from a Sherlock Holmes
point of view.

40mA is allowed for a maximum of 50ns (!).

I know, and 20mA max is recommended. But as I said, the current is limited
to some 15mA in several different ways. Under normal operating conditions
(i.e. with working feedback circuitry), it can't exceed 5mA. And the rest
of the original circuit is fine -- I stuck in another HCNR201, and it
worked perfectly right away. So no other components have failed, at least
not in any permanent manner.

Might just be bad luck.. but as someone else suggested, this sounds
like mechanical damage of some kind to me if the diode appears
electrically sound. No deliberate modification of the package such as
bending of leads?

Nope, I just bent the legs slightly inwards for normal assembly, as is usual
with DIP cases. And I'm pretty certain that I didn't overheat it either
during soldering.

Could the package be cracked at the leadframe or elsewhere due to
inadequate support of the PCB?

The PCB is quite small (5x8cm, or 2x3 inches approx.) and very well
supported. Also, the opto-isolator case doesn't show any cracks or other
damage.
But I guess I'll try to crack it open, although I seriously doubt if I can
find anything -- if only because cracking it open will certainly disturb
anything blocking the light path anyway.

But you could then fire up the LED again, pulse it with something, take
a photodiode from your parts bins, hook it up to the scope and see if
stuff is received.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

 

[Bob Larter]

Richard Rasker wrote:

Hi all,

I'm using an HCNR201 opto-isolator device in a galvanically isolated linear
signal transfer application. I have about two dozen of these in a not
completely unimportant application aboard sea ships, in a rather hostile
environment: the machine room, with heat (>40 degrees centigrade),
vibration and moisture. For these reasons, I designed the whole thing to be
very, very robust, and for a year or so, all was fine.

Recently, however, I got a complaint that one of those devices had failed,
and a bit of research showed that the opto-isolator was the cause: from the
outside, the LED still behaves like a LED diode, but none of the two photo
diodes produce any output -- which strongly suggests that the LED doesn't
produce any IR output any more.

Now I'm a bit puzzled by this, as the whole input circuitry is designed in
such a way that the opto-isolator LED would be among one of the very last
components to break down in case of a voltage spike or such -- there are
zener diodes, low-ohm SMD resistors and an SMD opamp which would blow
first, and under no conditions, should the total LED current be able to
exceed 15mA (with 40mA absolute maximum rating). Destructive testing with a
circuit here confirmed this: I managed to blow up a handful of parts --
twice -- but never the opto-isolator. Overvoltage, reverse voltage -- it's
all handled the way I designed it.

So my question: is this a simple case of "bad luck", or are there other ways
a LED in an opto-isolator may fail in this weird way (current OK, yet no
light)?

Sounds to me like you were simply unlucky & opto failed by itself.


--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

 

[Bob Larter]

Richard Rasker wrote:

Joerg wrote:

Richard Rasker wrote:
Joerg wrote:

Richard Rasker wrote:
Hi all,

I'm using an HCNR201 opto-isolator device in a galvanically isolated
linear signal transfer application.
...

So my question: is this a simple case of "bad luck", or are there
other
ways a LED in an opto-isolator may fail in this weird way (current OK,
yet no light)?

Both LED and photodiodes can die upon rather small reverse voltage
spikes. Without seeing your circuit it's hard to say which one is
vulnerable. I also don't know abs max because this miserable Acrobat
Reader crashed on that particular datasheet when scrolling.
(I use xpfd; works very fast, never crashes)

Yeah, I have to get something better than this dreaded Adobe stuff.
Foxit doesn't work with all docs but maybe xpdf does.

Um, I'm afraid xpdf is of little or no use to you -- it's Linux only, and
from your reference to Foxit I surmise you're running Windows.

You can run xpdf in Windows under Cygwin, although it's probably too
much effort to install Cygwin solely for a PDF viewer. OTOH, Cygwin is
fantastic if you want to be able to use Unix tools under Windows.

I usually have good experiences with those optical devices as well -- that's
also why I'm still a bit puzzled.

Ditto. I don't think I've ever seen a failed optocoupler.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

 

[Mark Zacharias]

"Richard Rasker" <spamtrap@linetec.nl> wrote in message
news:4aa937a8$0$733$7ade8c0d@textreader.nntp.internl.net...

Jan Panteltje wrote:

On a sunny day (Wed, 09 Sep 2009 20:55:49 +0200) it happened Richard
Rasker <spamtrap@linetec.nl> wrote in
4aa7fa35$0$725$7ade8c0d@textreader.nntp.internl.net>:

Jan Panteltje wrote:

On a sunny day (Wed, 09 Sep 2009 20:02:38 +0200) it happened Richard
Rasker <spamtrap@linetec.nl> wrote in
4aa7edbe$0$730$7ade8c0d@textreader.nntp.internl.net>:

So my question: is this a simple case of "bad luck", or are there other
ways a LED in an opto-isolator may fail in this weird way (current OK,
yet no light)?

Perhaps some part internal to the opto-isolater did break of due to
vibration and blocks the light path?
Did you open the defective one?

No, I didn't (yet) -- but isn't this very, very unlikely?

Richard Rasker

I could also imagine a bind wire coming lose of the photo transistor.
Before doing any destructive investigation, perhaps you could also check
the photo transistor.
Vbe and Vbc diodes, if it has the base on a pin, multimeter + on
collector, - on emittor, and feed the base from the collector via say
100k
Ohm. Should draw some current.

The HCNR201 isn't a normal opto-coupler; it doesn't have one photo
transistor, but two photo diodes, one of which is normally used in a
feedback circuit driving the LED. Both photo diodes behave the same, i.e.
they don't respond to any current I send through the LED.

Richard Rasker
--
http://www.linetec.nl

Well, they won't respond by producing a voltage; the forward resistance of
the diode lowers when the input LED is lit, I would think. Any voltage is
from the load circuit to which the photo diode output is connected, yes?

I admit I've only loosely been following this thread - and I'm used to the
traditional opto-isolator using an LED on the input side and a
phototransistor on the output side.

Mark Z.

 

[Richard Rasker]

Joerg wrote:

Richard Rasker wrote:
....
But I guess I'll try to crack it open, although I seriously doubt if I
can find anything -- if only because cracking it open will certainly
disturb anything blocking the light path anyway.


But you could then fire up the LED again, pulse it with something, take
a photodiode from your parts bins, hook it up to the scope and see if
stuff is received.

It's even simpler than that: any webcam and most cheaper digital cameras
can "see" an IR LED's output (this is also widely used to check remote
controls).

Richard Rasker
--
http://www.linetec.nl

 

[Richard Rasker]

Mark Zacharias wrote:

"Richard Rasker" <spamtrap@linetec.nl> wrote in message
news:4aa937a8$0$733$7ade8c0d@textreader.nntp.internl.net...

[snip]

The HCNR201 isn't a normal opto-coupler; it doesn't have one photo
transistor, but two photo diodes, one of which is normally used in a
feedback circuit driving the LED. Both photo diodes behave the same, i.e.
they don't respond to any current I send through the LED.

Well, they won't respond by producing a voltage; the forward resistance of
the diode lowers when the input LED is lit, I would think. Any voltage is
from the load circuit to which the photo diode output is connected, yes?

Actually, the photo diodes respond by generating a small current (typically
0.5% of the LED current); with a high-impedance voltmeter connected, this
results in a voltage in the order of one or two volts.

I admit I've only loosely been following this thread - and I'm used to the
traditional opto-isolator using an LED on the input side and a
phototransistor on the output side.

This particular device is suitable for galvanically isolated transfer of
analog signals with relatively high linearity. This is done by hooking up
one of the two photo diodes in the feedback of the LED drive circuit. With
both photo diodes closely matched, the other diode then delivers a current
accurately reflecting the input signal.

Richard Rasker
--
http://www.linetec.nl

 

[Spehro Pefhany]

On Thu, 10 Sep 2009 19:27:41 +0200, the renowned Richard Rasker
<spamtrap@linetec.nl> wrote:

Spehro Pefhany wrote:

On Wed, 09 Sep 2009 20:02:38 +0200, Richard Rasker
spamtrap@linetec.nl> wrote:

Hi all,

I'm using an HCNR201 opto-isolator device in a galvanically isolated
linear signal transfer application.

[snip mysterious breakdown]

So my question: is this a simple case of "bad luck", or are there other
ways a LED in an opto-isolator may fail in this weird way (current OK, yet
no light)?

When you say that the LED behaves like an LED diode, you mean that Vf
is just what you'd expect from figure 9 of the datasheet?

Hmm, I get a Vf of 1.3V @ 10mA If, so that's rather low, but still just
within specifications.

Typical is 1.3V at 10uA, and 1.55V at 10mA. That's a rather
significant deviation! (3 orders of magnitude in If)

Have you compared a fresh one?

AFAIUI, some modern LEDs (Double Heterstructure) etc. are more complex
than simple diodes and it seems possible this one has been damaged but
still retains diode-like characteristics.

material.eng.usm.my/stafhome/zainovia/EBB424e/LED3.ppt

I'd look a lot closer at the possibility of something bad getting in
to that LED (have you shunted it with a Si diode in reverse?).



Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

 

[Joerg]

Richard Rasker wrote:

Joerg wrote:

Richard Rasker wrote:
...
But I guess I'll try to crack it open, although I seriously doubt if I
can find anything -- if only because cracking it open will certainly
disturb anything blocking the light path anyway.

But you could then fire up the LED again, pulse it with something, take
a photodiode from your parts bins, hook it up to the scope and see if
stuff is received.

It's even simpler than that: any webcam and most cheaper digital cameras
can "see" an IR LED's output (this is also widely used to check remote
controls).

If you want to check remotes the photodiode on a scope is easier, heck,
even an LED will be sensitive enough as a photodiode. Then you can see
the transmission frequency and also all the codes. Makes it quite easy
to detect a short or bad button. I had that with the remote from an
80-year old. Pressed two different keys, same code came out. Opened it,
crud all over the board. Cleaned it, worked. "Oh, yeah, well, ahm, I
kinda did spill a can of V8 juice over it the other day".

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

 

[Nemo]

I can suggest a couple of things to check.

Opto couplers are complex mechanically, because they contain 2 different
types of component, which are then immersed in a resin which has to be
airtight, transparent to IR, have a thermal coefficient of expansion
matched to the rest of the device, and this is then covered by an opaque
layer to block out ambient IR. As a result of this compound construction
and all its failure modes, optos have a very high failure rate compared
to other components. You seem to be aware of this though as you mention
considering vibration, heat etc.

The manufacturers sometimes overlook one of these factors. There were
some cases of optocouplers a few years back (which used white plastic
rather than black bodies, I think) which didn't seem to do the last
function. They responded to ambient light / IR too. You'd think it was a
basic property of the device to NOT respond to this, but perhaps they
moved manufacture to a new factory and didn't emphasise the importance
of the resin type in the training. Or perhaps they were fake components.
If nothing else explains the problem, it may be worth checking the
supply chain to determine the provenance of these (expensive) devices.

The manufacturer's QA department should be able to comment on the batch
numbers written on the opto - a photo would probably be best if
possible, as this may give them other clues if they ARE fake like "all
our optos are in a different colour body".

When I had similar trouble a few years ago I asked the manufacturer's QA
department for failure rate information. It took a little while but they
produced it, which is when we realised how awful cheap optos were, and
switched to high quality ones (like the type you're using!). One of the
failure modes mentioned in the report was, IIRC 20 years later,
something about the resin going opaque as it aged! You should be able to
confirm that if you have an original circuit around, you could compare
its opto output with that of a new opto.

Does the circuit work OK if you drop a new opto into this position?

When all is said and done though, it sounds like your customers are
getting a bit overexcited about a single failure. As the application is
not unimportant (I assume this means it is for the defence industry), it
is a good idea to kick off an investigation to learn from the event and
improve the circuit if possible; but statistically a single failure in
one year conveys too little information to extrapolate.
--
Nemo